> ## Documentation Index
> Fetch the complete documentation index at: https://docs.zerotwo.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Deep Research

> Conduct comprehensive multi-source research with ZeroTwo's AI-powered research system

## What is Deep Research?

Deep Research is ZeroTwo's advanced research tool that goes beyond simple web search. It uses a multi-agent system to conduct thorough, comprehensive research on any topic, gathering information from multiple sources and synthesizing it into detailed reports.

<Note>
  Deep Research is like having a team of research assistants working in parallel - each investigating different aspects of your topic and combining their findings into a comprehensive report.
</Note>

## When to Use Deep Research

<CardGroup cols={2}>
  <Card title="Academic Research" icon="graduation-cap">
    Literature reviews, topic exploration, research paper preparation
  </Card>

  <Card title="Market Analysis" icon="chart-line">
    Industry trends, competitor analysis, market opportunities
  </Card>

  <Card title="Decision Making" icon="scale-balanced">
    Gather comprehensive information before important decisions
  </Card>

  <Card title="Due Diligence" icon="magnifying-glass-chart">
    Company research, product evaluation, technology assessment
  </Card>
</CardGroup>

## How Deep Research Works

### Multi-Agent Research System

Deep Research uses multiple AI agents working together:

<Steps>
  <Step title="Query Planning">
    AI breaks your research question into multiple sub-queries and research angles
  </Step>

  <Step title="Parallel Research">
    Multiple agents search different sources simultaneously:

    * Academic papers and publications
    * News articles and reports
    * Industry blogs and analysis
    * Technical documentation
    * Expert opinions and forums
  </Step>

  <Step title="Information Gathering">
    Each agent:

    * Searches relevant sources
    * Extracts key information
    * Identifies important quotes and data
    * Notes source credibility
  </Step>

  <Step title="Synthesis & Analysis">
    AI combines findings from all agents:

    * Identifies patterns and themes
    * Resolves contradictions
    * Highlights key insights
    * Organizes information logically
  </Step>

  <Step title="Report Generation">
    Creates comprehensive report with:

    * Executive summary
    * Detailed findings
    * Citations and sources
    * Recommendations (if applicable)
  </Step>
</Steps>

## Starting a Deep Research Session

<Tabs>
  <Tab title="From Chat">
    **Enable Deep Research tool**:

    1. Click the **Deep Research** toggle in the top bar
    2. Ask your research question
    3. AI initiates research process
    4. Watch progress in real-time

    ```
    Research the current state of quantum computing 
    and its practical applications
    ```
  </Tab>

  <Tab title="With Parameters">
    **Specify research parameters**:

    ```
    Research: Impact of AI on healthcare

    Focus on:
    - Last 3 years
    - Peer-reviewed sources
    - Clinical trials
    - Expert opinions

    Depth: Comprehensive
    Format: Academic report
    ```

    Provides more control over research scope.
  </Tab>

  <Tab title="From Template">
    **Use research templates** (coming soon):

    * Market Research Template
    * Academic Literature Review
    * Competitive Analysis
    * Technology Assessment
    * Due Diligence Report

    Pre-configured parameters for common research types.
  </Tab>
</Tabs>

## Research Depth Options

Choose the depth based on your needs:

<Tabs>
  <Tab title="Quick Overview">
    **Time**: 2-3 minutes\
    **Sources**: 5-10\
    **Length**: 500-1000 words

    **Best for**:

    * Quick topic overview
    * Initial exploration
    * Time-sensitive needs
    * Simple questions

    **Example**:

    ```
    Quick overview of GraphQL vs REST APIs
    ```
  </Tab>

  <Tab title="Standard Research">
    **Time**: 5-7 minutes\
    **Sources**: 10-20\
    **Length**: 1500-3000 words

    **Best for**:

    * Balanced research (default)
    * Most use cases
    * Good depth and breadth
    * Practical decisions

    **Example**:

    ```
    Research best practices for React state management
    ```
  </Tab>

  <Tab title="Comprehensive">
    **Time**: 10-15 minutes\
    **Sources**: 20-40\
    **Length**: 3000-6000 words

    **Best for**:

    * In-depth analysis
    * Academic work
    * Critical decisions
    * Complex topics

    **Example**:

    ```
    Comprehensive analysis of sustainable energy solutions 
    for developing nations
    ```
  </Tab>

  <Tab title="Deep Dive">
    **Time**: 15-20 minutes\
    **Sources**: 40+\
    **Length**: 6000+ words

    **Best for**:

    * Exhaustive research
    * Publication-quality work
    * Strategic planning
    * Multi-faceted topics

    **Example**:

    ```
    Deep dive into the economic, social, and environmental 
    impacts of artificial intelligence
    ```
  </Tab>
</Tabs>

## Monitoring Research Progress

Watch your research unfold in real-time:

### Progress Indicators

<AccordionGroup>
  <Accordion title="Research Phase">
    **What you'll see**:

    * Current phase (Planning, Researching, Synthesizing, Writing)
    * Progress percentage
    * Estimated time remaining

    **Phases**:

    1. 📋 Planning (10%) - Breaking down query
    2. 🔍 Researching (60%) - Gathering information
    3. 🧠 Synthesizing (20%) - Analyzing findings
    4. ✍️ Writing (10%) - Creating report
  </Accordion>

  <Accordion title="Sources Found">
    **Real-time updates**:

    * Number of sources discovered
    * Source types (academic, news, technical, etc.)
    * Quality indicators

    **Example display**:

    ```
    Sources: 15 found
    ├─ Academic papers: 6
    ├─ News articles: 4
    ├─ Technical docs: 3
    └─ Expert blogs: 2
    ```
  </Accordion>

  <Accordion title="Key Findings">
    **As research progresses**:

    * Important facts discovered
    * Emerging themes
    * Notable contradictions
    * Expert quotes

    Gives you a preview of what's being found.
  </Accordion>

  <Accordion title="Agent Activity">
    **See agents at work**:

    * Agent 1: Searching academic databases
    * Agent 2: Analyzing recent news
    * Agent 3: Exploring technical documentation
    * Agent 4: Finding expert opinions

    Shows parallel research happening.
  </Accordion>
</AccordionGroup>

<Tip>
  You can pause research at any time and AI will provide a report based on findings so far.
</Tip>

## Understanding Research Reports

### Report Structure

Deep Research reports follow a clear structure:

<Steps>
  <Step title="Executive Summary">
    **Quick overview** (200-300 words):

    * Main findings
    * Key insights
    * Important conclusions
    * Quick recommendations

    Perfect for decision-makers who need the highlights.
  </Step>

  <Step title="Methodology">
    **Research approach**:

    * Sources consulted
    * Search strategies used
    * Timeframe covered
    * Limitations noted

    Ensures transparency and credibility.
  </Step>

  <Step title="Detailed Findings">
    **Main body** organized by themes:

    * Section headings for clarity
    * Supporting evidence and data
    * Quotes from experts
    * Statistical information
    * Case studies and examples

    Comprehensive information organized logically.
  </Step>

  <Step title="Analysis & Insights">
    **Interpretation**:

    * Pattern identification
    * Trend analysis
    * Implications
    * Critical evaluation

    Goes beyond facts to provide understanding.
  </Step>

  <Step title="Sources & Citations">
    **Full bibliography**:

    * All sources cited
    * Links to originals
    * Publication dates
    * Credibility indicators

    Verify and explore further.
  </Step>

  <Step title="Recommendations">
    **Actionable next steps** (when applicable):

    * Based on findings
    * Prioritized actions
    * Further research areas

    Helps you act on insights.
  </Step>
</Steps>

### Citation Format

Each claim includes proper citations:

```markdown theme={null}
The global AI market is projected to reach $407 billion 
by 2027 [1], with healthcare applications growing at 
41.4% annually [2].

[1] Fortune Business Insights - AI Market Report 2024
    https://example.com/ai-market-report

[2] McKinsey & Company - Healthcare AI Analysis
    https://example.com/healthcare-ai
```

<Info>
  All citations are clickable links to original sources for verification.
</Info>

## Working with Research Results

### Export Options

<Tabs>
  <Tab title="Markdown">
    **Best for**: Documentation, GitHub, note-taking

    * Clean, readable format
    * Preserves structure
    * Includes all citations
    * Easy to edit

    Click **Export** > **Markdown**
  </Tab>

  <Tab title="PDF">
    **Best for**: Presentations, printing, sharing

    * Professional formatting
    * Page numbers
    * Table of contents
    * Ready to share

    Click **Export** > **PDF**
  </Tab>

  <Tab title="Word Document">
    **Best for**: Editing, collaboration, submissions

    * Microsoft Word compatible
    * Editable format
    * Comments and tracking
    * Academic submissions

    Click **Export** > **Word (.docx)**
  </Tab>

  <Tab title="Copy to Canvas">
    **Best for**: Further AI collaboration

    * Edit with AI assistance
    * Continue working
    * Integrate with other content
    * Real-time collaboration

    Click **Open in Canvas**
  </Tab>
</Tabs>

### Share Research

<CardGroup cols={2}>
  <Card title="Public Link" icon="link">
    Generate shareable link for anyone
  </Card>

  <Card title="Team Share" icon="users">
    Share with specific team members
  </Card>

  <Card title="Project Archive" icon="folder">
    Save to project for team access
  </Card>

  <Card title="Email" icon="envelope">
    Send directly via email
  </Card>
</CardGroup>

## Advanced Features

### Follow-Up Research

Continue research with follow-up queries:

<Steps>
  <Step title="Ask Follow-Up">
    ```
    Can you dive deeper into the healthcare applications section?
    ```
  </Step>

  <Step title="Targeted Research">
    AI researches specifically that aspect
  </Step>

  <Step title="Integrated Results">
    New findings integrated into existing report
  </Step>
</Steps>

### Compare Research

Compare multiple research reports:

```
Compare findings from:
1. AI in Healthcare research
2. Machine Learning in Medicine research

Highlight overlaps and differences
```

### Time-Based Research

Research how topics evolve:

```
Research quantum computing developments:
- 2020-2021
- 2022-2023
- 2024-present

Show progression and trends
```

## Research Quality Indicators

### Source Credibility

Research reports include quality indicators:

<CardGroup cols={2}>
  <Card title="High Credibility" icon="shield-check">
    * Peer-reviewed journals
    * Government publications
    * Established institutions
    * Primary sources
  </Card>

  <Card title="Medium Credibility" icon="shield">
    * Industry publications
    * Reputable news outlets
    * Expert blogs
    * White papers
  </Card>

  <Card title="Lower Credibility" icon="shield-halved">
    * General blogs
    * Opinion pieces
    * Unverified sources
    * Social media
  </Card>

  <Card title="Mixed Sources" icon="scale-balanced">
    Reports use primarily high and medium credibility sources
  </Card>
</CardGroup>

### Consensus vs. Debate

Reports indicate agreement level:

* **Strong Consensus**: Multiple sources agree
* **General Agreement**: Most sources aligned
* **Mixed Views**: Diverse perspectives
* **Active Debate**: Significant disagreement

<Tip>
  Areas of debate are highlighted with multiple perspectives presented fairly.
</Tip>

## Best Practices

<AccordionGroup>
  <Accordion title="Craft Clear Research Questions">
    **Better research questions**:

    ❌ "Tell me about AI"\
    ✅ "What are the current practical applications of AI in small business operations?"

    ❌ "Climate change"\
    ✅ "What are the most effective carbon reduction strategies for urban areas?"

    **Tips**:

    * Be specific about scope
    * Define the aspect you care about
    * Mention context or constraints
    * Specify timeframe if relevant
  </Accordion>

  <Accordion title="Specify Your Needs">
    **Include parameters**:

    ```
    Research: Remote work productivity

    Focus:
    - Studies from 2020-2024
    - Sample size > 100
    - Peer-reviewed preferred
    - Include statistics

    Audience: Business executives
    ```

    Helps AI understand what matters to you.
  </Accordion>

  <Accordion title="Review and Verify">
    **Always verify critical information**:

    * Click through to original sources
    * Check publication dates
    * Consider source credibility
    * Look for primary sources
    * Cross-reference key facts

    Deep Research is powerful but verification is your responsibility.
  </Accordion>

  <Accordion title="Use Appropriate Depth">
    **Match depth to need**:

    * Quick decisions → Quick Overview
    * Standard work → Standard Research
    * Important decisions → Comprehensive
    * Academic/strategic → Deep Dive

    Don't over-research simple questions.
  </Accordion>

  <Accordion title="Iterate and Refine">
    **Research process**:

    1. Start with standard research
    2. Review initial findings
    3. Ask follow-up questions
    4. Request deeper dives into specific areas
    5. Build comprehensive understanding

    Iterative approach often better than single deep dive.
  </Accordion>
</AccordionGroup>

## Example Research Sessions

### Market Research

```
Research the electric vehicle market:
- Current market size and growth
- Major players and market share
- Technology trends
- Consumer adoption barriers
- Future projections 2025-2030

Depth: Comprehensive
Focus: Business decision making
```

### Academic Literature Review

```
Literature review: Neural network architectures 
for natural language processing (2020-2024)

Include:
- Key papers and citations
- Architecture evolution
- Performance comparisons
- Current state of the art

Format: Academic style with full citations
```

### Technology Evaluation

```
Evaluate: Kubernetes vs. Docker Swarm for 
microservices deployment

Compare:
- Scalability
- Ease of use
- Community support
- Cost implications
- Use cases

Make recommendation for 50-person engineering team
```

## Limitations & Considerations

<Warning>
  **Important limitations to understand**:

  1. **Recency**: Most current for information from past 2 years
  2. **Paywalls**: Cannot access subscription-only content
  3. **Bias**: Sources may have inherent biases
  4. **Completeness**: May miss niche or very recent sources
  5. **Interpretation**: AI analysis should be verified
</Warning>

<AccordionGroup>
  <Accordion title="What Deep Research Can Do">
    ✅ Find and synthesize information from many sources\
    ✅ Identify patterns and trends\
    ✅ Compare different perspectives\
    ✅ Create structured reports\
    ✅ Provide citations and references\
    ✅ Save hours of manual research
  </Accordion>

  <Accordion title="What Deep Research Cannot Do">
    ❌ Access paywalled content\
    ❌ Guarantee 100% accuracy\
    ❌ Replace human expert judgment\
    ❌ Provide real-time minute-by-minute updates\
    ❌ Access private databases\
    ❌ Conduct primary research (surveys, experiments)
  </Accordion>
</AccordionGroup>

## Pricing & Usage

### Free Tier

* 3 Deep Research sessions per month
* Standard depth only
* All report features
* Full citations

### Pro Tier (\$20/month)

* 20 Deep Research sessions per month
* All depth options
* Priority processing
* Export to all formats

### Team Tier (\$30/user/month)

* Unlimited Deep Research
* Shared research library
* Collaborative annotations
* Custom templates

<Info>
  Deep Research credits reset monthly. Unused credits don't roll over.
</Info>

## Next Steps

<CardGroup cols={2}>
  <Card title="Web Search" icon="magnifying-glass" href="/tools/web-search">
    For quick, single-source information
  </Card>

  <Card title="Canvas" icon="pen-to-square" href="/tools/canvas">
    Edit and refine research reports
  </Card>

  <Card title="Projects" icon="folder" href="/workspaces/create-workspace">
    Organize research in workspaces
  </Card>

  <Card title="Memory" icon="brain" href="/tools/memory">
    Save key research insights long-term
  </Card>
</CardGroup>

<Check>
  Deep Research transforms hours of manual research into minutes of AI-powered investigation!
</Check>